CN-116436492-B - Blind estimation method for frequency hopping signal parameters based on atomic characteristics

CN116436492BCN 116436492 BCN116436492 BCN 116436492BCN-116436492-B

Abstract

The invention discloses a frequency hopping signal parameter blind estimation method based on atomic characteristics, which comprises the steps of firstly carrying out windowing treatment, compression sampling and atomic characteristic extraction on frequency hopping signals, and roughly estimating hopping time according to the distribution rules of the atomic characteristics of the signals in different windows; and finally, estimating the frequency of each hopping signal by using an orthogonal matching tracking algorithm, and obtaining the hopping pattern of the whole sample by splicing each hopping signal. The invention can change the calculation complexity and the estimation precision by adjusting the search step length and the compression ratio so as to adapt to different signal-to-noise ratio environments, can effectively reduce the calculation complexity by increasing the search step length and the compression ratio in the high signal-to-noise ratio environment, and can effectively improve the estimation precision by reducing the search step length and the compression ratio in the low signal-to-noise ratio environment.

Inventors

LEI YINGKE
JIN HU
FENG HUI
CHEN XIANG
WANG WEI
WANG YICHUAN
ZHU WEIPENG
WANG JIN
WANG YOURUI
TENG FEI
PAN BISHENG
QIAN FENG
JIANG LI
ZHANG MENGBO

Assignees

中国人民解放军国防科技大学

Dates

Publication Date: 20260505
Application Date: 20230322

Claims (9)

1. The method for blind estimation of the frequency hopping signal parameters based on the atomic characteristics is characterized by comprising the following steps: s1, windowing and sliding window processing are carried out on a frequency hopping signal to obtain signals in different windows; S2, compression sampling is carried out on signals in different windows to obtain inner product value distribution of the signals in different windows; S3, extracting atomic characteristics from the inner product distribution, arranging the atomic characteristics according to time sequence to obtain a distribution rule, and estimating the time range of the jump moment based on the distribution rule; s4, constructing a block diagonalized Fourier orthogonal matrix, wherein boundary points of the matrix traverse time ranges according to a set search step length, and estimating hopping moments through changes of sparse coefficients so as to obtain frequency hopping periods and hopping speeds; S5, estimating the frequency of each hopping signal by using an orthogonal matching pursuit algorithm, and obtaining a hopping pattern by combining hopping moments.
2. The blind estimation method of frequency hopping signal parameters based on atomic characteristics according to claim 1, wherein in S1, windowing and sliding window processing are performed on the frequency hopping signal to obtain different intra-window signals, specifically comprising the following steps: Step S11, sampling the intercepted frequency hopping signal according to the Nyquist sampling theorem, wherein the sampling frequency is 2 times of the highest frequency in the signal; and step S12, windowing and sliding window processing are carried out on the signal sampling data, different data in the window are intercepted, the window length is smaller than the data quantity of any one-hop signal, and the sliding step length is set to be one quarter of the window length.
3. The method for blind estimation of frequency hopping signal parameters based on atomic characteristics according to claim 1, wherein in S2, compression sampling is performed on signals in different windows to obtain inner product value distribution of the signals in different windows, specifically as follows: Step S21, performing compressed sampling on the intercepted data in the window, namely multiplying the data in the window by a Gaussian random measurement matrix to obtain a compressed measured value of the data in the window, wherein the compressed measured value is shown in the following formula: y=Φx wherein y is a compression measured value, x is data in a window, and phi is a Gaussian random measurement matrix; Step S22, the compression measurement value y and the recovery matrix Θ are subjected to inner product, and an inner product value and a distribution condition thereof are obtained as follows: Θ=ΦΨ Ω=|<Θ,y>| Wherein ψ is a discrete Fourier transform basis, Θ is a recovery matrix, and Ω is an inner product value.
4. The blind estimation method of the hopping signal parameters based on the atomic characteristics according to claim 1, wherein the time range of the hopping time is estimated in S3, specifically as follows: Step S31, extracting a maximum value val 1 and a second maximum value val 2 in the inner product distribution, and obtaining an atomic characteristic according to the following formula: step S32, arranging the atomic characteristics according to the time sequence to obtain a distribution condition; And step S33, locking the position of the trough in the atomic characteristic distribution condition, wherein the data in the window corresponding to the position of the trough is the time range of the jump moment.
5. The blind estimation method of frequency hopping signal parameters based on atomic characteristics according to claim 1, wherein in S4, a block diagonalized fourier orthogonal matrix is constructed, and the boundary points of the matrix traverse a time range according to a set search step length, and the hopping time is estimated through the change of a sparse coefficient, so as to obtain a frequency hopping period and a frequency hopping speed, which are specifically as follows: In step S41, a block diagonalized fourier orthogonal matrix is constructed, and the window contains at most two-hop signals, so that the matrix is represented as two diagonalized fourier orthogonal matrices, and the size of the matrix is (window length×window length), as shown in the following formula: Step S42, obtaining a specific search range according to the characteristic that the jump point is located at the middle part of the data in the window at the trough, wherein the specific search range is as follows: wherein N is the window length, delta is the sliding step length; Step S43, moving the demarcation point K in the matrix according to the set search step length, traversing the specific search range Obtaining and recording a sparse coefficient of each movement by using an orthogonal matching pursuit algorithm; Step S44, locking the largest sparse coefficient, wherein the boundary point position corresponding to the sparse coefficient is the jump time t hop ; Step S45, the difference between two adjacent hopping moments is a frequency hopping period, and the inverse of the frequency hopping period is a frequency hopping speed, as shown in the following formula: Wherein T hop is a frequency hopping period, and V hop is a frequency hopping rate.
6. The blind estimation method of hopping signal parameters based on atomic characteristics according to claim 1, wherein in S5, the frequency of each hopping signal is estimated by using an orthogonal matching pursuit algorithm, and a hopping pattern is obtained by combining hopping moments, specifically as follows: Step S51, the frequency of each jump signal is obtained by using an orthogonal matching pursuit algorithm, and the calculation formula is as follows: pos=argmax(|<Θ,y>|) Wherein f s is the sampling rate, pos is the position corresponding to the maximum value in the inner product value distribution; And step S52, splicing each hopping signal according to the time sequence to obtain a frequency hopping pattern.
7. A device for blind estimation of frequency hopping signal parameters based on atomic characteristics, the device comprising: The windowing and sliding window processing module is used for windowing and sliding window processing of the frequency hopping signals to obtain signals in different windows; the compression sampling module is used for carrying out compression sampling on the signals in different windows to obtain inner product value distribution of the signals in different windows; The time range determining module is used for extracting atomic characteristics from the inner product value distribution, arranging the atomic characteristics according to the time sequence to obtain a distribution rule, and estimating the time range of the jump time based on the distribution rule; The hopping time determining module is used for constructing a block diagonalized Fourier orthogonal matrix, and dividing points of the matrix traverse time ranges according to set searching step length, and the hopping time is estimated through the change of the sparse coefficient so as to obtain a frequency hopping period and a frequency hopping speed; And the frequency hopping pattern determining module estimates the frequency of each hopping signal by using an orthogonal matching pursuit algorithm and obtains the frequency hopping pattern by combining the hopping time.
8. A mobile terminal comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the method for blind estimation of frequency hopping signal parameters based on atomic characteristics according to any one of claims 1 to 6 when executing the program.
9. A computer readable storage medium, on which a computer program is stored, characterized in that the program, when being executed by a processor, implements the steps of the method for blind estimation of frequency hopping signal parameters based on atomic characteristics as claimed in any one of claims 1 to 6.

Description

Blind estimation method for frequency hopping signal parameters based on atomic characteristics Technical Field The invention belongs to the technical field of communication countermeasure, and particularly relates to a frequency hopping signal parameter blind estimation method based on atomic characteristics. Background The frequency hopping communication is widely applied in the military field by virtue of excellent anti-interception and anti-interference capabilities, and the reconnaissance of the non-own frequency hopping communication is already an important content on an informatization battlefield along with the continuous upgrading of communication countermeasure. The blind estimation of the parameters plays a crucial role as a key link of the whole reconnaissance process, and the estimated parameters can be used for subsequent decryption or interference. Because the frequency hopping signal has frequency domain sparsity, the signal sampling amount can be obviously reduced by fully utilizing the characteristic, and therefore, in recent years, the compressed sensing theory is applied to the blind estimation of the frequency hopping signal parameters. Frequency hopping signal parameter blind estimation research based on compressed sensing is not few, but the estimation accuracy of the algorithms needs to be improved when the signal to noise ratio is lower than 0dB, and the calculation complexity of the algorithms is higher when the signal to noise ratio is higher than 0 dB. Disclosure of Invention The invention aims to provide a method for blind estimation of frequency hopping signal parameters based on atomic characteristics, so that frequency hopping patterns of non-own frequency hopping signals are obtained, the estimation accuracy is higher under the condition of low signal to noise ratio, and the calculation complexity is lower under the condition of high signal to noise ratio. The technical scheme for realizing the purpose of the invention is that the method for blind estimation of the frequency hopping signal parameters based on the atomic characteristics comprises the following steps: s1, windowing and sliding window processing are carried out on a frequency hopping signal to obtain signals in different windows; S2, compression sampling is carried out on signals in different windows to obtain inner product value distribution of the signals in different windows; S3, extracting atomic characteristics from the inner product distribution, arranging the atomic characteristics according to time sequence to obtain a distribution rule, and estimating the time range of the jump moment based on the distribution rule; s4, constructing a block diagonalized Fourier orthogonal matrix, wherein boundary points of the matrix traverse time ranges according to a set search step length, and estimating hopping moments through changes of sparse coefficients so as to obtain frequency hopping periods and hopping speeds; S5, estimating the frequency of each hopping signal by using an orthogonal matching pursuit algorithm, and obtaining a hopping pattern by combining hopping moments. Further, in S1, windowing and sliding window processing are performed on the frequency hopping signal to obtain signals in different windows, which specifically includes: Step S11, sampling the intercepted frequency hopping signal according to the Nyquist sampling theorem, wherein the sampling frequency is 2 times of the highest frequency in the signal; and step S12, windowing and sliding window processing are carried out on the signal sampling data, different data in the window are intercepted, the window length is smaller than the data quantity of any one-hop signal, and the sliding step length is set to be one quarter of the window length. Further, in S2, compression sampling is performed on the signals in different windows to obtain inner product value distribution of the signals in different windows, which is specifically as follows: Step S21, performing compressed sampling on the intercepted data in the window, namely multiplying the data in the window by a Gaussian random measurement matrix to obtain a compressed measured value of the data in the window, wherein the compressed measured value is shown in the following formula: y=Φx wherein y is a compression measured value, x is data in a window, and phi is a Gaussian random measurement matrix; Step S22, the compression measurement value y and the recovery matrix Θ are subjected to inner product, and an inner product value and a distribution condition thereof are obtained as follows: Θ=ΦΨ Ω=|<Θ,y>| Wherein ψ is a discrete Fourier transform basis, Θ is a recovery matrix, and Ω is an inner product value. Further, in S3, the time range in which the jump time is estimated is specifically as follows: Step S31, extracting a maximum value val 1 and a second maximum value val 2 in the inner product distribution, and obtaining an atomic characteristic according to the following formula: step